Model Boundary
Methodology
The StatFC prediction system is a structured data model. Large language models explain and draft after model output exists; they do not invent fixtures, odds, injuries, lineups, or match facts.
The goal is not certainty. It is to show where model output, market pricing, and uncertainty diverge.
Every report must carry
Data timestamp
Model version
Market-implied probability
Uncertainty note
Review record
01
Data Input
Fixtures, odds, injuries, lineups, and scores must come from APIs or verified sources.
02
Structured Model
The model outputs probability, confidence, market gap, and scenarios instead of article conclusions.
03
Explanation Draft
AI explains model output, structures report drafts, and summarizes changes.
04
Human Review
Paid publication checks facts, timestamps, opposing scenarios, and risk boundaries.
05
Update Log
Major data changes leave a timeline entry instead of silently overwriting older judgment.
06
Post-match Review
The archive keeps original judgment, market movement, result, and review notes.
Credibility Boundary
No outcome promises and no certainty-led or profit-led language.
Unverified data is not written as fact.
Reports must include uncertainty and opposing scenarios.